Journal of Asian Economics 15 (2004) 399413

Short communication

The effect of foreign capital and imports on

economic growth: further evidence from four
Asian countries (19701998)
Kanta Marwaha,*, Akbar Tavakolib
a
b

Carleton University, Ottawa, Canada

University of Ottawa, Ottawa, Canada

Received 30 December 2003; received in revised form 15 February 2004; accepted 28 February 2004

Abstract
We measure and analyze the impact of foreign direct investment (FDI) and imports on the
economic growth and productivity of four individual countries which were among the founding
members of the Association of Southeast Asian Nations (ASEAN). These four countries are
Indonesia, Malaysia, the Philippines, and Thailand. After briefly reviewing their economic patterns,
we estimate, for each of the four countries, a separate production functions by using foreign capital
and imports as two distinct factors among the factors of production. Our analysis is based on time
series annual data from early 1970s to 1998. The estimated production elasticities of foreign capital
range from 0.044 for Thailand to 0.086 for Malaysia. About one-fifth to one-fourth of the productivity
of total capital stock is generated by growth in FDI: 24.5% in Indonesia, 25.8% in Malaysia, 21.4% in
the Philippines, and 20.3% in Thailand. Similarly, the production elasticities of imports range from
0.226 for Indonesia to 0.428 for Thailand. Our results show that openness measured by FDI netinflows and total imports jointly generates 0.292 of each growth point in Indonesia, 0.529 in
Malaysia, 0.353 in the Philippines, and 0.472 in Thailand.
# 2004 Elsevier Inc. All rights reserved.
Keywords: Foreign direct investment; Imports; Growth and productivity; ASEAN countries

1. Introduction
That openness of an economy helps economic growth is a widely accepted proposition.
There are two dimensions of openness: free trade in goods and services and free
*
Corresponding author. Tel.: 1-613-520-2600x3757.
E-mail address: kmarwah@ccs.carleton.ca (K. Marwah).

1049-0078/$ see front matter # 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.asieco.2004.02.008

400

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

international capital flow. Earlier research on the subject was focused on the free-trade
dimension, especially on export-led growth.1 The basic argument put forward was that
whereas exports stimulate economic growth primarily from the demand side, they also
produce efficiency gains, by a way of global competition, on the supply side. Lately,
import-led growth has been more in focus, and faster growing developing countries have
experienced much activity emerging from importing. Import-led growth emphasizes the
process of modernization and transfer of advanced technology through acquisition of much
needed sophisticated capital and material. Basically, this argument hinges on the growth of
total factor productivity and is centered on the supply side. The burgeoning endogenousgrowth trade literature has further fortified the argument in favor of openness and trade
liberalization. For this literature suggests that trade contributes to economic growth largely
by opening access to intermediate inputs; by expanding diffusion of knowledge; by
amplifying a learning-by-doing process; and by expanding the global market size (Iscan,
1998).2
The other dimension of openness relates to the capital stock flowing freely across
international borders, and here the focus of research has been on the foreign direct
investment (FDI). Empirical evidence that FDI has made a positive contribution to the
economic growth of developing countries (DCs) has accumulated fast. Some recent
examples of case studies are Marwah and Klein (1998) for India, Li, Liu, and Rebelo
(1998) and Sun (1998) for China, Ramirez (2000) for Mexico, and Djankov and Hoekman
(2000) for Czech Enterprises. There also exist multi-country studies based mostly on crosssection or panel data. Balasubramanyan, Salisu, and Sapsford (1996), Blomstrom, Lipsey,
and Zegan (1994), Borensztien, de Gregorio, and Lee (1998), de Mello (1999), Jackman
(1982), Most and Van Den Berg, 1996, Nair-Reichert and Weinhold (2001), and Teboul and
Mouslier (2001) are a few examples. de Mello (1997) provides an annotated selective
survey of earlier studies.
In his well cited study, Jeffrey Sachs (2000) examined a catching-up growth process as
one important pattern of development for DCs. He linked the catching-up growth process
to the openness of the economy. In this process an economy with a lower level of
technology and income (the follower) narrows the income gap with the higher
technology and richer countries (the leader) through a process of technological diffusion
and capital flows from leader to follower (p. 581). Imports and liberalization of FDI are
two important elements that act as conduit for absorption of foreign technology by the
follower. Sachs examined 150 DCs and found that only 24 of these countries, all with
successful export-promotion policies and attracting large FDI inflows, had won the race in
absorbing technologies from abroad.3
In this paper we provide further evidence on how imports and FDI inflows have
contributed to the growth and productivity of four individual countries. These four
countriesnamely Indonesia, Malaysia, the Philippines, and Thailandare the founding
1

For earlier literature and survey of studies, see Brander (1992) and Edwards (1993).
Marwah and Klein (1996) show a strong impact of openness on growth through a trade-based entropy index.
Harrison (1996) finds the relationship between openness and economic growth weak.
3
Through export-promoting policies a developing economy is able to earn the foreign exchange necessary to
import technologies. The FDI inflows add to the foreign exchange resources as well as facilitating upgrading the
countrys technologies.
2

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

401

members of the Association of Southeast Asian Nations (ASEAN) and they also made
Sachss list of 24 successful DCs. Our choice of these four countries for individual
country analysis is further underscored by an earlier finding of Marwah and Klein (1996)
that the ASEAN region had surpassed during 19711989 the worlds average growth
rate by 2.72%. This conclusion was based on their analysis of the relationship between
the growth rate and a trade entropy construct.4 In this paper we estimate, for each
country, a separate production function using imports and foreign capital as two distinct
factors of production. The analysis is based on time series annual data from early 1970s
to 1998.
In Section 2, the economic patterns of the four selected countries are briefly highlighted.
The methodological production function is specified in Section 3, and the estimates are
presented in Section 4. The final Section 5 contains some concluding remarks.

2. A retrospect of economic patterns

The four Asian countriesIndonesia, Malaysia, the Philippines, and Thailand
selected for our analysis differ widely in size and availability of natural resources. Yet
these four countries have exhibited a certain degree of similarity in their patterns of
economic development and growth. All have export-oriented market-based economies; all
reflect common economic aspirations with much emphasis on economic growth; and all
have dualistic economies with more than 50% of their populations living in an agricultural
and rural sector (Tongzon, 1998, p. 13). Moreover, they all have pursued economic
development and growth through international trade and investment, and over the past
three decades have changed their policies considerably. By and large, their export
promotion policies were initiated in the 1970s, but these policies got launched on a full
scale only during the late 1980s (Sakurai, 1995, p. 175). In an attempt to attract more
investment, they all made their rules and regulations of FDI less restrictive. The main
sources of their FDI inflows are the United States, Japan, and the European Union (EU). In
the first half of 1990s, nine East Asian countries, namely China, Hong Kong, Indonesia,
(Republic of South) Korea, Malaysia, the Philippines, Singapore, Taiwan and Thailand,
had attracted together a total of US$ 206.4 billion in FDI flows, out of which 24.3% had
gone to Indonesia, Malaysia, the Philippines, and Thailand.
Since the late 1970s, the foreign economic policy has changed moderately in Thailand
but significantly in Indonesia, Malaysia, and the Philippines. In Indonesia, restrictions and
barriers to foreign investments were either modified or removed with trade liberalization in
1986; in the Philippines, the rules and regulations on FDI were relaxed by the Foreign
Investment Act of 1991. In Malaysia, since 1986 the role of government intervention has
gradually declined and many new incentives to attract FDI have been initiated (Tongzon,
1998, p. 151).
Overall, changes in investment and foreign trade policies of the four countries seem to
have been conditioned by how well their economies did perform. Growth rates from the
4
The concept of entropy is borrowed from the information theory (see Theil, 1971). It was used as a measure
of openness of the economy.

402

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

Table 1
Some comparative macro-statistics (ratios & growth rates) (%)
Variables

Period

Growth rates
Indonesia

Malaysia

Philippines

Thailand

GDP

19721998
19811998

6.0
5.7

6.7
7.0

2.7
2.2

7.2
7.7

Investment (I)

19721998

10.1

9.0

3.4

8.7

FDI

19721998
19811998
19721998
19761998
19721998
19731998
19761998
19761998
19761998
19761998
19701998
19701998


23.8
9.1
2.9
13.5
15.1
5.2
9.4
12.4
8.6
9.3
10.8

9.8
14.3
10.8
3.2
12.8
13.1
3.3
8.7
8.9
12.0
0.3
4.1


22.1
6.3
3.0
7.1
15.3
0.2
2.0
10.5
5.7
7.1
12.0

8.7
13.9
9.8
2.5
11.6
13.8
5.6
8.7
11.9
8.8
1.5
6.1

30.2
4.3
60.4

21.6
0.8
29.9

29.7
1.1
31.3

Imports (M)
Employment
Kd
Kf
GDP/labor
Kd/labor
Kf/labor
Imports/labor
Exchange rate
Inflation rate

Ratios
I/GDP
FDI/GDP
M/GDP

19701998
19701998
19701998

23.4
0.8
22.9

Source: calculated in the study. All variables have been defined in the appendix.

For some years, FDI net-flows were negative.

For period 19811997.

early 1970s in selected key macroeconomic variables and the associated entities of each of
the four countries are portrayed in Table 1.
2.1. Indonesia
Indonesia has a rich endowment of natural resources and a relatively large domestic
market. It is an oil exporter. In 1998, its population was about 204 million people. Prior to
early the 1980s Indonesia pursued chiefly an import-substitution policy, keeping in view a
large enough size of its domestic market. An export-promotion policy initiated at the
beginning of the 1980s by measures such as deregulation of local content restrictions was
pursued more aggressively in the late 1980s. FDI was gradually liberalized in the 1980s
when the restrictions on foreign investors were slowly relaxed (Sakurai, 1995, p. 188).
As shown in Table 1, the ratios of Indonesias imports to GDP (22.9%) and net FDI
inflows to GDP (0.8%) are relatively low. However, both imports and foreign capital
intensities of labor seem to have accelerated fast over time, by annual rates of 8.6 and 9.4%,
respectively. The growth rate of labor productivity of Indonesia (5.2%) is somewhat lower
in comparison to the rate of Thailand (5.6%), but it is substantially higher than the rate of
Malaysia (3.3%). Among the four countries, Indonesia ranks third by the GDP growth rate

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

403

and by the investment/GDP ratio, but it ranks first by the growth rate of real domestic
investment (10.1%).
On inflation rate, Indonesia experienced an annual rate of 10.8% during 19701998,
which is much higher than the rates experienced by Malaysia and Thailand. Furthermore, in
this period, the value of Indonesias Rupiah depreciated the most in comparisons with
Malaysias Ringgit, Philippines Peso and Thailands Baht. It fell almost by 9.3% per annum.
2.2. Malaysia
Overall, the Malaysian economy has performed much better than the other three
economies. It showed somewhat superior growth rates of GDP and employment, lower
rate of inflation and a relatively less fluctuating exchange rate. Even though Malaysia, with
a population of 22 million in 1998, is considered small in size, it enjoys a high income level
because of its rich endowment of natural resources. It, too, is an oil exporter like Indonesia.
From Table 1, its real GDP has grown quite fast, at a rate of 6.7% per annum, over 1972
1998. The growth rate of its domestic investment was 9%. The domestic investment has
constituted on an average 30% of GDP, the highest rate among the four countries.
Initially, in the 1960s and the early 1970s, Malaysia followed an import-substitution
policy. Import liberalization and export-promotion policies were adopted later in the
1970s. Free trade zones played an important role in promoting Malaysian exports (Sakurai,
1995, p. 189). The FDI inflows limited by the size of its domestic market stagnated till the
mid-1980s, but then started to expand steadily once the new incentives were put into place.
As a result of foreign trade and investment liberalization policies, the average ratio of net
FDI inflows to GDP rose significantly during 19701998. Its value of 4.3% is a high value
compared to corresponding ratios in Indonesia (0.8%), the Philippines (0.8%), and
Thailand (1.1%). In Indonesia and the Philippines, FDI was probably discouraged by
dictatorship and insurgencies. The Malaysian share of FDI net inflows of DCs has also
increased dramatically. Since 1970, Malaysia has remained as one of the top 12 recipients
of FDI among DCs (IFC, 1997). Its real imports have also grown fast at a 10.8% rate per
annum and its import intensity has gone up even faster at an annual rate of 12%. The share
of its imports in GDP on an average was 60% during 19701998 (Table 1). When compared
with the corresponding rates of the other three countries these rates turn out to be the
highest.
The inflation rate in Malaysia has been significantly lower compared to other countries.
The consumer price index (CPI) went up at a rate of 4.1% annually during 19701998. The
CPI registered a growth rate of 3.1% for the last two decades, but it changed even at a
slower (2.7%) rate during the 1980s. At the same time, during the 1970s, the value of its
currency appreciated by a rate of 3.3% annually, whereas in the previous two decades it had
depreciated by 1.5% annually. Overall, the Malaysian Ringgit appreciated over the entire
period by 0.3%. It fluctuated more moderately than the currencies of other three countries.
2.3. The Philippines
An import-substitution policy implemented in the early 1950s in the Philippines
continued until the 1970s. At the same time, the Philippines devalued its Peso, repeatedly,

404

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

and also imposed restrictions and high tariffs on imports until the late 1970s (Sakurai,
1995, p. 189). It was only after the adoption of economic reforms in the 1990s that import
restrictions were largely lifted and tariff rates were reduced. However, FDI inflows to the
Philippines continued to decline so long as the country failed to abandon its restrictive
import-substitution policies.
Overall performance of the Philippines economy has been rather poor compared to the
other three countries. Its per capita GNP, US$ 770, in 1998 remained almost at the same
low level that was achieved nearly two decades earlier. At the same time, during 1970
1998, its CPI rose annually by 12% (Table 1); that was the fastest among the four countries.
While both the monetary growth rate and the inflation rate fluctuated widely in 19831986,
the interest rate rose steeply. It is believed that the high interest rate squeezed out the private
investment and consequently widened the gap between the ratios of private investment to
GDP of the Philippines and Thailand (Mutoh, 1995, p. 89). The GNP growth rate of the
Philippines declined continuously during the first half of the 1980s, so much so that it
became negative in 1984 and 1985.
The Philippines Peso depreciated annually on an average rate of 7.1% during 1970
1998. It was devalued on two occasions, in 1970 and 1984. But these devaluations seemed
to help neither in expanding exports nor in stimulating capital inflows. Apparently, both
devalued currency and the double digit inflation had a damaging effect on the Philippines
economy which was already struggling to catch up with the other three countries. Nor did
the Philippines have good infrastructure for FDI. The Philippines growth rate of real GDP
and of average labor productivity has been low. The domestic capital intensity has grown
more slowly than in the other three countries. In short, the Philippines economic
performance has lagged far behind its neighbors, and consequently, the economic gap
between the Philippines economy and the other three has been widening over time. Indeed
Ferdinand Marcos, crony capitalism, and uprising that the country had to deal with have
left their marks.
2.4. Thailand
Before the late 1960s, Thailands economy was a poor agricultural economy, very
much in line with (or even lagging behind) the Philippines economy (Kawagoe &
Sekiguchi, 1995, p. 14). But as indicated above, there has now emerged a big gap between
the two economies. After the late 1960s, and until the recent financial crises of mid-1997,
Thailand had a stable economic growth, low inflation rate and a manageable level of
external debt. Its import-substitution policy was initiated in the 1960s and it continued
until the mid-1970s; its export-promotion policy was adopted in the late 1970s and the
early 1980s (Sakurai, 1995, p. 190). Thailands import-substitution strategy was seriously
affected by two oil crises in the 1970s. Since the end of the 1980s, Thailand has achieved a
high rate of economic growth, a boom in FDI and expanded exports of manufactured
goods.
On comparison with the other three countries, Thailands economic performance
ranks next to Malaysia. Its 5.6% growth rate of average labor productivity and 7.2%
growth rate of GDP were highest among the four countries. High ratios of its domestic
investment, net FDI, and imports to GDP rank this country in the second position (Table 1).

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

405

Since 1970, along with Malaysia, Thailand has been one of the top 12 recipients of FDI
among DCs (IFC, 1997). Moreover, Thailands factor intensity of domestic capital, that is
domestic capitallabor ratio, has grown in parallel with Malaysias but its factor intensity
of foreign capital has been expanding even faster. In fact, an extensive program of
economic deregulation in its international transactions has been put in place since
1990. The restrictions on external payments related to current account transactions were
almost completely liberalized in May 1991, and the restrictions on external capital
accounts were greatly relaxed enabling Thailand to accept the Article (8) obligations
under the IMF (Mutoh, 1995, p. 73). The inflation rate (CPI) of 6.1% in Thailand, on an
average during 19701998, has been higher than the corresponding rate of 4.1% in
Malaysia.
Overall, in 1979 and the early 1980s, DCs were faced with different international crises.
The oil price increase of 1979, the rise in the international interest rates that started in
1979, the recession in the major world economies in the early 1980s, and the international
debt crisis that erupted in 1982, they all created a very unfavorable international climate for
these countries (Jansen, 1995, p. 194). Indonesia, Malaysia, the Philippines, and Thailand
were all affected deeply and directly during this period. Especially, their currencies were
tied too closely to the US dollar. Consequently, during the appreciation period of the dollar,
they priced themselves out of world markets and the worsening balance-of-payment
situation generated crisis.
The Asian financial crisis of 19971998 exacted a heavy toll of the economies of these
countries. In order to cope with the crisis, Indonesia and Thailand were forced to embark on
the IMF-mandated program by committing to float the exchange rate and tighten both
monetary and fiscal policies. Malaysia response to the crisis was different. It did not call in
IMF. Instead it followed an independent route. It imposed sweeping controls on capital
flows out of the country, lowered the interest rate, and revalued the Ringgit upward. Kaplan
and Rodrik (2001) examined the question, Did the Malaysian gamble pay off? They
found some evidence that Malaysian capital controls did allow a speedier recovery than
would have been possible via the orthodox/IMF route.

3. Model specification
Most available empirical studies aimed at measuring an impact of FDI and imports on a
developing economy are based on some version of a neo-classical growth model. In these
studies either a homogenous CobbDouglas production function, or its parallel growth rate
accounting equation derived from the production function is estimated directly. In this
paper, using a relatively long period annual data, we estimate and test a basic multifactor
CobbDouglas production function and its two generalized versions for Indonesia,
Malaysia, the Philippines and Thailand. That is, in each case, we test three alternative,
but closely related, specifications. These specifications are CobbDouglas (CD) production function, transcendental (TR) production function and the CES generalized Cobb
Douglas (CES:CD) function. Our final choice, individually for each case, reported in the
paper, is determined by statistical results. We estimate the production functions by using
capital stock, labor and imports as factors of production. There are also some auxiliary

406

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

variables such as time trend and structural dummies. But we make a distinction between
domestic and foreign capital stocks as two separate factors. Since we include imports as a
factor of production, on the left-hand side, we use gross output (GX), that is GDP plus
imports as the dependent variable. The use of gross output has been suggested strongly by
Klein (1992) for consistency with an adding-up or product-exhausting theorem. That is, if
every factor is paid according to its marginal productivity, then the payments to factors
must equal output inclusive of profit. (Also, see Marwah & Klein, 1998.) If we assume the
functional relationship between output and factor inputs as:
GX FKd ; Kf ; L; M; D; m

(1)

where Kd is domestic capital stock, Kf is foreign direct capital stock, L labor, M imports,
D surrogate for any auxiliary variable(s) and m stochastic random error, then an accounting
growth equation for analyzing output and productivity over time is:
d ln GX
@ ln F d ln Kd @ ln F d ln Kf @ ln F d ln L @ ln F d ln M

dt
@ ln Kd dt
@ ln Kf dt
@ ln L dt
@ ln M dt
@ ln F d ln D

@ ln D dt

(2)

By (2), the rate of growth of output is expressed as a weighted sum of the rates of growth
of factor inputs, where the factor-weights are their respective production elasticities. Thus
the first term on the right hand side measures the contribution to growth of domestic capital,
the second of foreign capital, the third of labor and the fourth of imports. We estimate the
production elasticities by experimenting with following three functional forms of the
production function, but finally we choose for each country the one which yields the best
statistical results.
A. CobbDouglas production function (CD):

a1
a2
b P
GX
Kd
Kf
M
A
e i di Di  m1
(3)
L
L
L
L
B. Transcendental production function (TR):

a1
a2
b
P
GX
Kd
Kf
M
A
eg1 Kd =Lg2 Kf =Lg3 M=L i di Di  m2
L
L
L
L

(4)

C. CES generalization of CobbDouglas production function (CES:CD):

a1
a2
b
P
2
2
2
GX
Kd
Kf
M
A
eg1 ln Kd ln L g2 ln Kf ln L g3 ln Mln L i di Di  m3
L
L
L
L
(5)
In (3)(5), a1, a2 and b are >0, gi and di may be positive or negative, and Dis are dummy
variables. Furthermore, if all gis are equal to zero in (4) and (5), both specifications
converge to CobbDouglas function (3). An assumption of constant returns to scale is
subsumed in the above functional forms. Whereas the CobbDouglas function yields
constant values of production elasticities and unitary value of elasticity of substitution, the

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

407

underlying production technologies of (4) and (5) allow for both non-unitary elasticities of
substitution and variable elasticities of production. (See, for example, Intriligator, Bodkin,
& Hsiao; 1996, chapter 8, pp. 293297.)
Defining production elasticity (Z) of GX with respect to factor input ( ) as
Z @GX=@ (( )/GX), the production-elasticity functions from these specifications
as shown in Marwah and Klein (1998) are:
From A, Eq. (3),
ZKd a1 ;

ZKf a2 ;

ZM b;

ZL 1  ZKd  ZKf  ZM

From B, Eq. (4),




Kd
Kf
ZKd a1 g1
; ZKf a2 g2
;
L
L


M
; ZL 1  ZKd  ZKf  ZM
ZM b g3
L
From C, Eq. (5),




Kd
Kf
ZKd a1 2g1 ln
; ZKf a2 2g2 ln
;
L
L


M
; ZL 1  ZKd  ZKf  ZKM
ZM b 2g3
L

4. The estimates
Our estimates are based on annual data for the sample period 19701998. Data are drawn
from two main sources: (a) the International Monetary Fund (IMF, various issues) and the
Asian Development Bank (ADB, various issues). All value entities, unless specified
otherwise, are defined in terms of national currencies, and are measured in real terms
by using 1995 prices. For Indonesia, the values are expressed in billions of Rupiah, for
Malaysia, in millions of Ringgit, for the Philippines, in billions of Peso, and for Thailand,
in billions of Baht. (For detailed information on the definition and construction variables,
see Appendix A.)
The best estimates of the production functions, in logarithmic transformation, for the
four countries are presented below. The estimates are obtained by ordinary least squares
(OLS). The numbers within parentheses below the coefficients are t-ratios. The coefficient

2 ), DurbinWatson ratio (d), the mean
of determination adjusted for degrees of freedom (R
of the dependent variable, and the standard error of estimate (S.E.) have been listed with
each equation. Three dummy variables denote structural shifts over three distinct time
periods. These are:
D1 1 for 19701983, 0 elsewhere,
D2 1 for 19841988, 0 elsewhere,
D3 1 for 19891998, 0 elsewhere, and
T time trend.

408

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

Indonesia (19761998, N 23):








GX
Kd
Kf
M
ln
0:203 ln
0:066 ln
0:226 ln
L
L
L
L
3:5
3:0
4:8
1:53 D1 1:42 D2 1:43 D3
20:3

14:5

12:6

(6)

2 0:976, d 1:6[
AR(2)].
Mean 1:608, S:E: 0:01528, R
Note: AR(2) denotes auto-regressive error of order 2, corrected using ML (Gauss
Newton) method.
Malaysia (19701998, N 29):








GX
Kd
Kf
Kf
ln
0:247 ln
 0:061 ln
0:251 D3 ln
L
L
L
L
9:1
4:5
5:3






M
Kf
M
 0:02
0:385 ln
0:005
L
L
L
21:5
4:7
3:4
1:72 D1 1:71 D2 1:29 D3
31:4

29:5

10:1

(7)

0:999, d 2:0.
Mean 3:3307, S:E: 0:00105, R
The Philippines (19731998, N 26):








GX
Kd
Kf
M
ln
1:12 1:125 ln
0:066 ln
0:287 ln
L
L
L
L
2:4
3:7
2:6
9:3
0:184ln Kd  ln L2  0:02T
3:3

7:2

2 0:927, d 2:0.
Mean 2:3237, S:E: 0:01419, R
Thailand (19721998, N 27):








GX
Kd
Kf
M
ln
0:84 0:145 ln
0:044 ln
0:295 ln
L
L
L
L
5:6
6:3
2:4
6:2




Kd
M
0:17
5:026
L
L
0:9
2:9

(8)

(9)

2 0:999, d 2:1.
Mean 2:3442, S:E: 0:0050, R
It may be noted that the CD production function describes best the production
technology in Indonesia, TR, in Malaysia and Thailand, and truncated CES:CD, augmented by time trend, in the Philippines. Except for one single coefficient in Eq. (9), all
other coefficients of Eqs. (6)(9), are statistically significant and are robust with a priori
^ underscore a positive and strong contribution
expected signs. The estimated values of b, b,
of imports to average labor productivity in each country. Furthermore, the estimated
(non-zero) value of g3 appears only in equations for Malaysia and Thailand. Its positive
values, in both cases, imply increasing marginal returns to imports. Evidently, this result is
consistent with the fact that the ratio of imports to GDP is high for both countries.
Importantly, a clear distinction emerges between the impact of domestic and FDI capital
stock for each country. Both ^
a1 and ^
a2 are positive and statistically significant. Specifically,
with an exception of Malaysia, ^
a2 is very similar across the other three countries. For

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

409

Malaysia, the direction of effect seem to have shifted in the 1990s, from negative to positive.
It is shown by the coefficient of multiplicative dummy variable. In other words, the effect of
Kf on labor productivity has been varying over time, but the positive effect since 1989 dominates the negative effect of the earlier years. For other three countries, the effect is normal.
The dummy variables seem to capture some shifts over time in total factor productivity.
Apparently, these shifts are caused by changes in trade and investment policies in Indonesia
and Malaysia; however, a negative time trend is revealed in total factor productivity in the
Philippines.
Parenthetically, it may be mentioned that statistical properties of the residuals, namely
normality, heteroscedasticity and serial correlation were tested and the residuals were
found to be stationary except for Indonesia, where DurbinWatson d-statistic was weak.
The estimates were then obtained after correcting for autocorrelation as indicated above.
The estimated equations also passed the functional form test.
For Indonesia, partial production elasticities are given by the coefficient of each
respective factor in Eq. (6) and are thus constant, but for the other three countries, they
vary from point to point. For easy comparison, the variable elasticities have been evaluated
at mean functional points of respective Eqs. (7)(9). The elasticity coefficients are
summarized in Table 2.
The closeness of production elasticities of both domestic and FDI capital stock across
four countries is quite striking. For domestic capital, it varies from 0.173 to 0.247, and for
FDI, from 0.044 to 0.086. With respect to labor and imports, the production elasticities are
very similar in Indonesia and the Philippines on the one hand, and of Malaysia and
Thailand on the other. In Indonesia and the Philippines, the production elasticity of labor is
higher than of imports, and the opposite is true for Malaysia and Thailand.
Finally, and parenthetically, it may be noted that we did test the super consistent
(robustness) properties of OLS estimators by applying the cointegration test for the
relationship among the main variables in a specified model (Verbeek, 2000, p. 282; Zhang,
2001). Two steps are followed. First, the augmented DickeyFuller (ADF) test is applied to
confirm that all variables, used in the main equation are rendered stationary in second
differences (D2) at a 5% level of significance, or they are integrated of order two, I(2), as
shown below in Table 3. Since nearly all basic variables used in the model are integrated of
the same order, the error term in the equation is integrated of zero order, I(0). Second, the
cointegration property of each equation was checked by applying Johansens LR test,
which included a constant term and no time trend with the VAR lag lengths set at 2. This
test showed that the equation is cointegrated for each country at 5% critical level. A priori
Table 2
Implied (partial) production elasticities (Z( ))
Z( )

Indonesia

Malaysia

The Philippines

Thailand

ZKd
ZKf
Z(L)
Z(M)

0.203
0.066
0.505
0.226

0.247
0.086a
0.224
0.443

0.242
0.066
0.405
0.287

0.173
0.044
0.355
0.428

All variable (partial) production elasticities are computed at mean points.

a
For period 19891998.

410

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

Table 3
Test-Results for the Unit Root of Variables
Variable D2( )

Indonesia
19761998

Malaysia
19721998

Philippines
19721998

Thailand
19731998

D2 ln (GX/L)
D2 ln (Kd/L)
D2 ln(Kf/L)
D2 ln(M/L)
ADF critical
value (5%)

4.22 (4.13)
3.05 (E)
4.48 (5.61)
5.86 (5.61)
3.0294 (3.6746)

4.00 (4.06)
4.24 (3.77)
7.92 (8.70)
4.90 (4.96)
2.9850 (3.60271)

3.96 (3.86)
5.69 (5.65)
3.61 (3.94)
6.81 (6.55)
2.9907 (3.6119)

4.26 (4.60)
3.03 (E)
E
4.74 (4.87)
2.9970 (3.219)

Statistics are based on with intercept and no trend, and within parentheses, with trend and intercept. When the
absolute value of statistic listed in the Table is greater than the absolute value of ADF statistic (the last row), the
unit root for the corresponding variable is rejected. The letter E denotes not possible to reject the unit root.

expected signs for each individual coefficient were also confirmed. Overall, and importantly, it is legitimate to interpret our regressions as structural equations.

5. Conclusions
In this paper, the impact of foreign direct investment and imports is analyzed for four
ASEAN countries: Indonesia, Malaysia, the Philippines and Thailand. Based on the
production elasticities evaluated at their functional mean points over a relatively long
period of 19701998, results are best summarized, individually for each country, by its
growth-accounting Eq. (2) transformed in statistical form. That is,
Indonesia:
d ln GX
d ln Kd
d ln Kf
d ln L
d ln M
0:203
0:226
0:066
0:505
dt
dt
dt
dt
dt
Malaysia:
d ln GX
d ln Kd
d ln Kf
d ln L
d ln M
0:247
0:443
0:086
0:224
dt
dt
dt
dt
dt
The Philippines:
d ln GX
d ln Kd
d ln Kf
d ln L
d ln M
 0:242
0:287
0:066
0:405
dt
dt
dt
dt
dt
Thailand:

(10)

(11)

(12)

d ln GX
d ln Kd
d ln Kf
d ln L
d ln M
0:173
0:428
(13)
0:044
0:355
dt
dt
dt
dt
dt
Evidently, both foreign capital and imports have marked effect on economic growth of
Indonesia, Malaysia, the Philippines and Thailand. The message that emerges is clear: for
every 1% growth point, growth of domestic and foreign capital stock generates 0.269 in
Indonesia, 0.333 in Malaysia, 0.308 in the Philippines and 0.217 in Thailand. Furthermore,
of every 1% growth point generated by the growth of total capital stock, growth in foreign

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

411

capital stock accounts for 24.5% in Indonesia, 25.8% in Malaysia, 21.4% in the Philippines, and 20.3% in Thailand.5 Although, the production elasticities of foreign capital stock
range from 0.044 for Thailand to 0.086 for Malaysia, the closeness of relative contribution
of growth in foreign capital to growth in domestic capital in a range of 2025% is a
powerful finding. That one-fifth to one-fourth of the productivity of total capital stock is
generated by growth in FDI is not an insignificant contribution. At the same time, every 1%
growth in imports generates 0.226 of each growth point in Indonesia, 0.443 in Malaysia,
0.287 in the Philippines, and 0.428 in Thailand. That openness has contributed substantially to the productivity and economic growth of ASEAN countries is clear by our results.
To sum up, openness to big industries, measured by FDI net-inflows and total imports,
generates 0.292 of each growth point in Indonesia, 0.529 in Malaysia, 0.353 in the
Philippines, and 0.472 in Thailand.
Acknowledgements
We are thankful to Lawrence R. Klein and Ronald G. Bodkin for very helpful comments.
All remaining errors are ours.
Appendix A. Definitions of variables
Note: The national currencies are: Rupiah in Indonesia, Ringgit in Malaysia, Peso in the
Philippines, and Baht in Thailand. All values for Indonesia, the Philippines and Thailand
are expressed in billion, but for Malaysia in million, of national currency units. Exchange
rate is defined as national currency per US$ at the end of the period (IMF series, rf).
GDP:
GX:
Kd:

Kf:

Gross domestic product (X) in 1995 national currency units. Nominal values
are converted to real values by using GDP implicit deflator, 1995: 1.00.
Gross domestic product plus imports (GDP M).
Domestic capital stock in 1995 national currency units. Domestic capital stock is
calculated by accumulating gross domestic investment (gross fixed capital
formation), using 5% depreciation rate and 1969 as the base year. That is,
Kd t 1  0:05Kd t1 Id t , where Id is gross fixed domestic capital
formation. Kd 0 Id 1969t . Nominal values of gross domestic investment are
converted to real values by using domestic investment implicit deflator, 1995: 1.00.
Foreign capital stock in 1995 national currency units. It is calculated by
accumulating net-FDI flows and using 5% depreciation rate. The base year
is 1970, except for the Philippines, which is 1973.
Net-inflows of Foreign Direct Investment (FDI), defined as FDI inflows minus
FDI outflows, are obtained in US dollars. The values are converted appropriately
in national currency units, and in 1995 values, by using a proper exchange rate
and implicit investment price deflator.
For the Philippines, the data are available for 19731998.

ZKf
ZKd ZKf .

412

M:
L:
T:

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

Imports in 1995 national currency units. The implicit import price deflator
is used to convert the nominal values to real terms.
Employment, thousands of persons. For Indonesia, total employment is
available only from 1976.
Chronological time.

References
Asian Development Bank (various issues). Key indicators of developing and Pacific countries. Oxford University
Press.
Balasubramanyan, V. N., Salisu, M., & Sapsford, D. (1996). Foreign direct investment and growth in EP and IS
countries. The Economic Journal, 106, 92105.
Blomstrom, M., Lipsey, R. E., & Zegan, M. (1994). What explains the growth of developing countries? In W. E.
Baumal, R. R. Nelson, & E. N. Wolff (Eds.), Convergence of productivity: Cross-national studies and
historical evidence. Oxford: Oxford University Press.
Borensztien, E., de Gregorio, J., & Lee, J. W. (1998). How does foreign direct investment affect economic
growth? Journal of International Economics, 45, 115135.
Brander, J. (1992). Comparative economic growth: evidence and interpretation. Canadian Journal of Economics,
25(4), 792818.
de Mello, L.R., Jr. (1997). Foreign direct investment in developing countries: A selective survey. Journal of
Development Studies, 34(1), 134.
de Mello, L.R., Jr. (1999). Foreign direct investment-led growth: Evidence from time series and panel data.
Oxford Economic Papers, 51(1), 133151.
Djankov, S., & Hoekman (2000). Foreign investment and productivity growth in Czech enterprises. The World
Bank Economic Review, 14 (1), 4964.
Edwards, S. (1993). Openness, trade liberalization, and growth in developing countries. Journal of Economic
Literature, 31(3), 13581393.
Harrison, A. (1996). Openness and growth: A time series, cross-country analysis for developing countries.
Journal of Development Economics, 48(2), 419447.
IFC (1997). Foreign direct investment, Report No. 5. Washington, DC: World Bank.
International Monetary Fund (various issues) International financial statistics yearbook. Washington, DC: IMF.
Intriligator, M. D., Bodkin, R. G., & Hsiao, C. (1996). Econometric models, techniques, and applications
(2nd ed.). NJ: Prentice Hall.
Iscan, T. (1998). Trade liberalization and productivity: A panel study of the Mexican manufacturing industry.
The Journal of Development Studies, 34(5), 123148.
Jackman, R. W. (1982). Dependence of foreign investment and economic growth in the third world. World
Politics, 34, 175196.
Jansen, K. (1995). The macroeconomic effects of direct foreign investment: The case of Thailand. World
Development, 23(2), 193210.
Kaplan, E., & Rodrik, D. (2001, February). Did the Malaysian capital controls work? NBER Working Paper No.
8142, Cambridge, MA.
Kawagoe, T., & Sekiguchi, S. (Eds.) (1995). East Asian economies: Transformation and challenge. Singapore:
Institute of Southeast Asian Studies.
Klein, L. R. (1992). Restructuring of the world economy. In B. G. Hickman (Ed.), International productivity &
competitiveness (pp. 3348). Oxford: Oxford University Press.
Li, H., Liu, Z., & Rebelo, I. (1998). Testing the neoclassical theory of economic growth: Evidence from Chinese
provinces. Economics of Planning, 31(2/3), 117132.
Marwah, K., & Klein, L. R. (1996). What are some prospects for Indias joining the Asian growth process? In R.
Hooley, et al. (Eds.), Asia Pacific economic cooperation: Theory and practice, Research in Asian economic
studies 7 (Part B) (pp. 419439). JAI Press.

K. Marwah, A. Tavakoli / Journal of Asian Economics 15 (2004) 399413

413

Marwah, K., & Klein, L. R. (1998). Economic growth and productivity gains from capital inflow: Some evidence
for India. Journal of Quantitative Economics, 14(1), 81108.
Most, S. J., & Van Den Berg, H. (1996). Growth in Africa: Does the source of investment financing matter?
Applied Economics, 28(9), 14271433.
Mutoh, T. (1995). Macroeconomic performance of developing economies: A comparison of Korea, Thailand and
the Philippines since 1966. In T. Kawagoe & S. Sekiguchi (Eds.), East Asian economies: Transformation and
challenge (pp. 72103). Singapore: Institute of Southeast Asian Studies.
Nair-Reichert, U., & Weinhold, D. (2001). Causality tests for cross-country panels: A new look at FDI and
economic growth in developing countries. Oxford Bulletin of Economics and Statistics, 63(2), 153172.
Ramirez, M. D. (2000). Foreign direct investment in Mexico: A cointegration analysis. Journal of Development
Studies, 37(1), 138162.
Sachs, J. D. (2000). Globalization and patterns of economic development. Weltwirtschaftliches, Archiv, 136(4),
579600.
Sakurai, M. (1995). East Asian trade and investment policies. In T. Kawagoe & S. Sekiguchi (Eds.), East Asian
economies: Transformation and challenge (pp. 172197). Singapore: Institute of Southeast Asian Studies.
Sun, H. (1998). Macroeconomic impact of direct foreign investment in China: 197996. The World Economy,
21(5), 675694.
Teboul, R., & Mouslier, E. (2001). Foreign aid and economic growth: The case of the countries south of the
mediterranean. Applied Economics Letters, 8(3), : 187190.
Theil, H. (1971). Principles of econometrics. New York: John Wiley.
Tongzon, J. L. (1998). The economies of Southeast Asia: The growth and development of ASEAN economies.
MA: Edward Elgar.
Verbeek, M. (2000). A guide to modern econometrics. New York: John Wiley & Sons, Ltd.
Zhang, K. H. (2001). Does foreign direct investment promote economic growth? Evidence from East Asia and
Latin America. Contemporary Economic Policy, 19(2), 175185.